Vision-based tactile sensors (VBTSs) are a promising technology for robots, providing them with dense signals that can be translated into an understanding of normal and shear load, contact region, texture classification, and more. However, existing VBTS tactile surfaces make use of silicone gels, which provide high sensitivity but easily deteriorate from loading and surface wear. We propose that polyurethane rubber, used for high-load applications like shoe soles, rubber wheels, and industrial gaskets, may provide improved physical gel resilience, potentially at the cost of sensitivity. To compare the resilience and sensitivity of silicone and polyurethane VBTS gels, we propose a series of standard evaluation benchmarking protocols. Our resilience tests assess sensor durability across normal loading, shear loading, and abrasion. For sensitivity, we introduce model-free assessments of force and spatial sensitivity to directly measure the physical capabilities of each gel without effects introduced from data and model quality. Finally, we include a bottle cap loosening and tightening demonstration as an example where polyurethane gels provide an advantage over their silicone counterparts.

Interaction between learner and feedback-giver has come into focus recently for post-training of Large Language Models (LLMs), through the use of reward models that judge the appropriateness of a model's response. In this paper, we investigate whether Dialogue Games -- goal-directed and rule-governed activities driven predominantly by verbal actions -- can also serve as a source of feedback signals for learning. We introduce Playpen, an environment for off- and online learning through Dialogue Game self-play, and investigate a representative set of post-training methods: supervised fine-tuning; direct alignment (DPO); and reinforcement learning with GRPO. We experiment with post-training a small LLM (Llama-3.1-8B-Instruct), evaluating performance on unseen instances of training games as well as unseen games, and on standard benchmarks. We find that imitation learning through SFT improves performance on unseen instances, but negatively impacts other skills, while interactive learning with GRPO shows balanced improvements without loss of skills. We release the framework and the baseline training setups to foster research in the promising new direction of learning in (synthetic) interaction.

Recently, several approaches successfully demonstrated that weight-sharing Neural Architecture Search (NAS) can effectively explore a search space of elastic low-rank adapters (LoRA), allowing the parameter-efficient fine-tuning (PEFT) and compression of large language models. In this paper, we introduce a novel approach called Shears, demonstrating how the integration of cost-effective sparsity and a proposed Neural Low-rank adapter Search (NLS) algorithm can further improve the efficiency of PEFT approaches. Results demonstrate the benefits of Shears compared to other methods, reaching high sparsity levels while improving or with little drop in accuracy, utilizing a single GPU for a pair of hours.

Weak gravitational lensing of distant galaxies provides a powerful probe of dark energy. The aim of this study is to investigate the application of convolutional neural networks (CNNs) to precision shear estimation. In particular, using a shallow CNN, we explore the impact of point spread function (PSF) misestimation and `galaxy population bias' (including `distribution bias' and `morphology bias'), focusing on the accuracy requirements of next generation surveys. We simulate a population of noisy disk and elliptical galaxies and adopt a PSF that is representative of a Euclid-like survey. We quantify the accuracy achieved by the CNN assuming a linear relationship between the estimated and true shears and measure the multiplicative ($m$) and additive ($c$) biases. We make use of an unconventional loss function to mitigate the effects of noise bias and measure $m$ and $c$ when we use either: (i) an incorrect galaxy ellipticity distribution or size-magnitude relation, or the wrong ratio of morphological types, to describe the population of galaxies (distribution bias); (ii) an incorrect galaxy light profile (morphology bias); or (iii) a PSF with size or ellipticity offset from its true value (PSF misestimation). We compare our results to the Euclid requirements on the knowledge of the PSF model shape and size. Finally, we outline further work to build on the promising potential of CNNs in precision shear estimation.

Most vision-based tactile sensors use elastomer deformation to infer tactile information, which can not sense some modalities, like temperature. As an important part of human tactile perception, temperature sensing can help robots better interact with the environment. In this work, we propose a novel multimodal vision-based tactile sensor, SATac, which can simultaneously perceive information of temperature, pressure, and shear. SATac utilizes thermoluminescence of strontium aluminate (SA) to sense a wide range of temperatures with exceptional resolution. Additionally, the pressure and shear can also be perceived by analyzing Voronoi diagram. A series of experiments are conducted to verify the performance of our proposed sensor. We also discuss the possible application scenarios and demonstrate how SATac could benefit robot perception capabilities.

Tactile servoing is an important technique because it enables robots to manipulate objects with precision and accuracy while adapting to changes in their environments in real-time. One approach for tactile servo control with high-resolution soft tactile sensors is to estimate the contact pose relative to an object surface using a convolutional neural network (CNN) for use as a feedback signal. In this paper, we investigate how the surface pose estimation model can be extended to include shear, and utilize these combined pose-and-shear models to develop a tactile robotic system that can be programmed for diverse non-prehensile manipulation tasks, such as object tracking, surface following, single-arm object pushing and dual-arm object pushing. In doing this, two technical challenges had to be overcome. Firstly, the use of tactile data that includes shear-induced slippage can lead to error-prone estimates unsuitable for accurate control, and so we modified the CNN into a Gaussian-density neural network and used a discriminative Bayesian filter to improve the predictions with a state dynamics model that utilizes the robot kinematics. Secondly, to achieve smooth robot motion in 3D space while interacting with objects, we used SE(3) velocity-based servo control, which required re-deriving the Bayesian filter update equations using Lie group theory, as many standard assumptions do not hold for state variables defined on non-Euclidean manifolds. In future, we believe that pose and shear-based tactile servoing will enable many object manipulation tasks and the fully-dexterous utilization of multi-fingered tactile robot hands. Video: https://www.youtube.com/watch?v=xVs4hd34ek0

When Sam Altman, the chief executive of OpenAI, took to the stage in San Francisco nine days ago he hinted at another significant development in the world of artificial intelligence. "Four times now in the history of OpenAI, the most recent time was just in the last couple weeks, I've gotten to be in the room, when we sort of push the veil of ignorance back and the frontier of discovery forward, and getting to do that is the professional honour of a lifetime," he told the Asia-Pacific Economic Cooperation (Apec) summit. Given that he leads the company behind ChatGPT – a chatbot that has transformed the debate around AI – this was a tantalising comment. And a major event in AI did occur the next day – Altman was fired. OpenAI's board announced on Friday 17 November that it had sacked the 38-year-old for failing to be "consistently candid in his communications" with its members, without giving further details about the alleged breaches of trust.

Sam Altman is set to make a return as chief executive of OpenAI after the ChatGPT developer said it had "reached an agreement in principle" for his reinstatement. The San Francisco-based company made the announcement after days of corporate drama in the wake of Altman's surprise sacking on Friday. Nearly all of OpenAI's 750-strong workforce had threatened to quit unless the board overseeing the business brought back Altman and then quit immediately afterwards. As part of the agreement reached overnight, the deal includes a new-look board led by Bret Taylor, the former co-CEO of software firm Salesforce. It will include Larry Summers, the former US treasury secretary, and Adam D'Angelo, the tech entrepreneur and current board member who played a role in Altman's firing.

Sam Altman is returning to OpenAI as CEO after his firing five days ago launched the company onto one of the wildest rollercoaster rides in tech history, the company announced in post on X. Former president Greg Brockman, who resigned on Friday in protest, will also return, The Verge's sources say. The original board has been disbanded and replaced by a new, temporary three-man board with Bret Taylor (chair), Larry Summers and original board member Adam D'Angelo. The agreement has been struck "in principal," and must still be approved by all parties. The only job of the initial board will be to vet and appoint a permanent board with up to 9 members that will resent OpenAI's governance.

Even though it seemed that former OpenAI CEO Sam Altman would lead a new AI research division at Microsoft, he might still get his old job back. According to Bloomberg, the OpenAI board -- which caused chaos at the company when it fired Altman on Friday -- has reopened discussions with the former chief executive regarding his possible reinstatement. The talks are said to involve board member (and Quora CEO) Adam D'Angelo as well as OpenAI investors, some of whom have been pushing for Altman's return. According to the report, board members "largely refused to engage" with Altman until Monday, so these latest talks are said to be a significant development. Meanwhile, Kevin Scott, Microsoft's chief technology officer, said that his company will match the compensation OpenAI workers are currently receiving if they jump ship.